Les Cramer and His Super Speech Machine

Compressed speech can be used to teach the blind, relay instructions to Air Force pilots, take phone orders and uncover psychological problems-all without sounding like Donald Duck.

Les Cramer is a dreamer. Six years ago, while president of his own oil company in Needham, he suddenly became interested in human communications systems. Now, at age forty, Cramer has developed a technique of recording the spoken word that has revolutionized everything from education for the blind to the frozen food industry.

In his small office in Larsen Hall, cluttered with graphs, tapes and electronic equipment, H. Leslie Cramer seems more the mad scientist than the Ed School Ph.D. candidate. It is here that Cramer perfected a process to compress speech by deleting small word segments. Previous experimenters had attempted to speed speech by retaping it at a faster rate -- producing only unintelligible Donald Duck gabblings. But Cramer's process enables the listener to hear and comprehend up to 1000 words per minute.

The technique itself is relatively uncomplicated -- bits of syllables are arbitrarily discarded by a special electromagnetic speech compressor machine. Cramer has found, through extensive experimentation, that the ideal "discard interval" is 14 milliseconds, or only 14/1000 of every second of recorded speech. Yet the result is good comprehension at rates up to 3 times as fast as normal conversation.

The practical uses of compressed speech are diverse and often complex. Its most immediate implications are in the field of education for the blind. When reading Braille, a blind person can achieve a maximum rate of 125 words per minute, provided he is skilled enough to use both hands. The average Braille reading rate is only 90 w.p.m. With Cramer's speeded speech program, however, blind children could listen to compressed speech lectures through earphones in languagelab situations at anywhere from 475 to 600 w.p.m. In the future, new techniques may raise intelligibility rates to 1000 w.p.m. A speeded speech program has already been instituted on an experimental basis at a school for the blind in Winnetka, Illinois. The Library of Congress has also expressed interest in Cramer's findings for its talking-book programs for the blind.

Cramer is currently working on a machine which would make it possible for a blind student to regulate the speed of the tape he is listening to. A brighter student could progress according to his own ability, unhindered by the rest of the class.

The blind, however, comprise only a half of one percent of the nation's handicapped. There are those who are not legally blind, but who nonetheless have poor eyesight and experience difficulty reading. For these people, Cramer's technique makes talking book programs more practical. WGBH in Boston is considering broadcasting novels -- best sellers and popular mysteries -- in compressed speech two days each week.

Cramer has been perfecting his present system for a year. He first became interested in human communication back in 1960, during a political controversy in his home town. "If a politician stood up at a bridge meeting and cried 'Wolf!' it was amazing to see how quickly the public responded, while they reacted hardly at all to authoritative information on the printed page," Cramer observed. He concluded that auditory information was far more effective than written.

As he began to probe the matter, he discovered other information to support his theory. "The Russians have conducted very interesting experiments with dogs. For instance, they trained an animal to expect bread at the sound of a bell, but meat if a light were flashed instead. After a while, when both signals were employed at the same time, the animal invariably went for the bread, even though as a rule dogs prefer meat."

Cramer didn't use dogs in his experiments, but rather 160 Cliffies over a one-year period. The girls listened through earphones to speech varying in speed from 450 to 930 w.p.m., and were asked to write down whatever they could understand from the tapes. Cramer didn't use any Harvard men in his testing. "I wanted to see what could be achieved under optimum conditions," he said. "Women have been shown to have a consistently higher verbal ability than men," who, he added quickly, "are usually better at math."

Out of this experiment with Cliffies came a flood of new ideas, and more original uses for compressed speech. As the tapes' discard intervals increased, or, as larger bits of words were deleted, the girls would write down what they thought they had heard, which often had no relation to what had actually been spoken on the tape. This occurrence was not strange in itself; however, what was curious was that a number of girls would write the same "wild" answer, often a proper name. Intrigued, Cramer has begun work on "auditory Rorschach tests," in which he deletes such large segments of the taped speech that whole syllables are missing. The subject must then reconstruct entire sentences, in an ambiguous context that closely parallels a Rorschach inkblot.

Cramer has devised an alternate test to discover psychological problems. A student will, through earphones, hear the same story told in both ears. At a crucial point in the narrative, the left ear will hear one word, while the right will hear another word of equal length. For example, in a story about a man and a woman, the important pair of words might be "he killed/kissed her." It has been proven that the student's mind will subconsciously choose to record only one of the words; which one is chosen becomes the psychologist's problem.

A variation of this technique may be practicable in a new type of occupational test, Cramer feels: one that will increase a student's chance of discovering his true interests, divorced from external pressures that may have been exerted upon him. Instead of hearing a story through his earphones, the student will hear a succession of pairs of words -- for instance, stethoscope/habeas corpus -- and his subconscious will choose the one he will remember. In current aptitude tests, in which the student is given a set of questions and multiple-choice answers and will have time to consider the ramifications of his pick, the test's value may be ruined.

Certain branches of the Government have picked up techniques like Cramer's. The Air Force, for instance, will employ his findings to delay speech in one ear for pilots and control tower operators who must communicate through noise interference. Cramer has discovered that a listener tends, as he hears another person speak to latch on to certain tonal qualities in the speaker's voice. As he listens, he will be able to hear and comprehend what the person is saying even through noise interference. With this in mind, instead of speeding speech to save time, Cramer has developed a process to "delay" it for the sake of intelligibility and comprehension. Through experience, a pilot will hear a sentence on the left, say, slightly before he hears the same thing on the right. The reinforced tonal qualities will enable the pilot to understand instructions despite distracting noises. The delay itself is so miniscule that even an echo can be heard.

The FBI and CIA have also adopted the speeded speech concept to cut the time on listening to hours of taped material, such as meetings, routine information, hearings and interviews.

Condensed Gripes

Industry, too, has begun to benefit from speeded speech as a time- and labor-saving device. For example, when you call the phone company to register a complaint, chances are that you're talking to a machine, who will record your gripe and play it back to the operator in speeded speech form. One frozen food company has put a computer to work taking orders. Once the order has been restructured in compressed speech form, all the preliminary formalities have been worked out, leaving the operator time to fill two or three times as many orders as usual.

Among the many future uses that Cramer envisions for speeded speech is "auditory speed reading." In the Harvard Remedial Reading course the addition of a sound track has been proposed to complement the movies